Michael Tyrimos is the Founder/CEO of Capacitor Partners, an operations and technology consultancy that transforms large-scale enterprises.
getty
Walk into the boardroom of an oil company, a mining firm or a government agency and you’ll find two things. A polished deck full of digital transformation ambitions. And, if you ask the right questions, a quiet admission that most of the pilots from the last three years never actually scaled.
This is the defining paradox of digital transformation in 2026. Every industry claims to be doing it. Budgets are approved, consultants are hired, and programs are launched. Yet a survey of 3,186 CIOs and technology executives across 88 countries found that fewer than half of all digital initiatives meet or exceed their business outcome targets. And that average conceals something worse: Certain industries are failing at rates that would concern any executive who looked closely at the numbers.
Construction is a useful place to start because it rarely gets mentioned in these conversations. The sector employs 7% of the world’s working-age population, yet labor productivity has remained essentially flat for two decades while manufacturing productivity nearly doubled. McKinsey’s industry digitization research has long identified it as one of the least digitized sectors in the economy, just above agriculture.
It doesn’t improve much from there. The public sector and energy companies have evolved the least among the industries measured, according to the BCG Digital Acceleration Index, which assessed organizations across 27 countries. Most still run siloed, business unit-driven operating models, which is precisely the structure that makes scaling any digital initiative nearly impossible.
In mining, Deloitte’s most recent sector analysis is unusually direct: Slow digital adoption is described as a critical vulnerability as global competition and sustainability pressures continue rising.
The hardest number of all comes from McKinsey’s Global Survey on digital transformations. In oil and gas, automotive, infrastructure and pharmaceuticals, transformation success rates fall to between 4% and 11%. Less than 1 in 10 efforts in these sectors deliver. That’s not underperformance. That’s a structural problem.
The contrast with leading industries is not subtle. Companies at the front of AI adoption achieve 1.7 times the revenue growth and 3.6 times the three-year total shareholder return of organizations left behind, according to BCG research surveying 1,250 executives across nine industries in 2025. Only 5% of companies globally qualify as genuinely future-ready; the remaining 35% are scaling, and the 60% that remain are either stagnating or scaling too slowly to close the distance.
The gap is also accelerating. McKinsey’s analysis of over 1,000 companies globally shows that the spread in digital and AI maturity between leaders and laggards grew by 60% from 2016-2019 to 2020-2022. What was a manageable deficit is now structural, and the compounding nature of digital advantage means every year of inaction makes recovery harder.
Look across every sector in the data above, and the same three problems keep appearing, regardless of geography or company size.
The first is infrastructure that was never built for this. Oil fields, mines, hospital systems and government IT estates were designed for operational reliability, not data interoperability. Retrofitting them is expensive and disruptive, which means most organizations delay it, and the longer they delay it, the further behind they fall.
The second is a workforce culture that doesn’t reward experimentation. In industries where a failed process can trigger a regulatory investigation or cost hundreds of millions, people learn quickly not to take risks they don’t have to take. That instinct protects the business day to day. It also makes it nearly impossible to run the kind of fast, iterative, fail-and-learn cycles that digital transformation demands.
The third is what I’d call the pilot trap. Every one of these industries has dozens of transformation initiatives running at any given moment. The pilots work. Results look strong in a contained environment. And then nothing happens. The organization defaults back to what it knows, the team moves on to the next initiative, and the insight generated never makes it into how the business actually operates. After a while, the transformation program becomes a permanent fixture that everyone references, and nobody takes seriously.
What the data makes clear, and what I’ve observed consistently, is that the technology is seldom the real obstacle. The obstacle is that changing how a business operates is genuinely hard, and these industries have spent decades building cultures, incentive structures and operating models that were specifically designed to resist that kind of change.
Cost is the pressure point most organizations underestimate. More than 90% of CIOs told Gartner that managing cost limits their ability to capture value from AI, and that productivity gains vary by role and experience level, not simply by whether tools have been deployed. Giving people access to technology and building the conditions for them to use it effectively are two completely different things.
In asset-heavy industries, technology investment without a corresponding shift in the operating model has no measurable impact on business performance, a pattern documented in Deloitte’s analysis of the energy and industrials sectors. The operating model has to move alongside the tooling. One without the other consistently underdelivers.
The industries currently behind are not incapable of change. Companies within oil and gas, mining and the public sector that have transformed successfully share one consistent trait: a leadership team that treated this as a business redesign challenge, not a technology procurement decision. The tools were always available. What took time was the will to change how the organization actually worked around them.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。